Bulletin

sTi

rtrn

BULLETIN

OF

^ysfcal &
A Pplied Sci.
Serials

The Terrestrial Ele&ric
Observatory

OF
FERNANDO SANFORD

Palo Alto, California

Volume VI

Observations upon EleAric Repulsion between Insulated Bodies and Elec-
tric Attradion between Insulated and Uninsulated Bodies Which Are Caused
by Variations in the Distribution of the Earth's Eledric Charge Induced
by the Electrostatic Indu&ion of the Negative Charge of the Sun, and a
Discussion of the Origin of the Twelve -hour Barometric Wave and Its
Relation to the Atmospheric Potential Gradient

Palo Alto, California
March 1930

BULLETIN

OF

The Terrestrial Ele&ric
Observatory

OF
FERNANDO SANFORD

Palo Alto, California

Volume VI

Observations upon Eleclric Repulsion between Insulated Bodies and Elec-
tric Attradion between Insulated and Uninsulated Bodies Which Are Caused
by Variations in the Distribution of the Earth's Eleclxic Charge Induced
by the Ele&rostatic Induction of the Negative Charge of the Sun, and a
Discussion of the Origin of the Twelve -hour Barometric Wave and Its
Relation to the Atmospheric Potential Gradient

Palo Alto, California
March 1930

CONTENTS

PAGE

Brief review of the work on the electrostatic induction of the sun . 5
On a method of determining the distribution of the earth's surface

charge 7

Results of experimental tests 9

Table showing mean daily electrometer deflections . 11, 12, 13, 14, 15

Lunar electrostatic induction 15

Some bearings of the observations upon electrical theory .... 19
Daily variations in atmospheric pressure and the distribution of the

earth's surface charge 19

Attempted explanation of the twelve-hour barometric wave ... 27

Relation of barometric wave to earth-currents 29

Atmospheric potential gradient and barometric pressure .... 29

OBSERVATIONS UPON ELECTRIC REPULSION BETWEEN
INSULATED BODIES AND ELECTRIC ATTRACTION BE-
TWEEN INSULATED AND UNINSULATED BODIES WHICH
ARE CAUSED BY VARIATIONS IN THE DISTRIBUTION OF
THE EARTH'S ELECTRIC CHARGE INDUCED BY THE ELEC-
TROSTATIC INDUCTION OF THE NEGATIVE CHARGE OF
THE SUN, AND A DISCUSSION OF THE ORIGIN OF THE
TWELVE-HOUR BAROMETRIC WAVE AND ITS RELATION
TO THE ATMOSPHERIC POTENTIAL GRADIENT

Brief review of the work on the electrostatic induction of the sun. —
As this is probably the last Bulletin which will be published from this
observatory, it seems advisable to give a short review of the work which
has been done here and the reason why its separate publication was neces-
sary.

In 1911 the present writer published a paper in the Leland Stanford
Junior University Series entitled A Physical Theory of Electrification.
In this paper an attempt was made to show how all the known phenomena
of static electricity could be explained without making any assumptions
different from those generally believed at that time except the assumption
that the earth carried a very great charge of negative electricity.

It had already been shown by Hale that the regions around sun-spots
were the seats of enormous charges of negative electricity, and the later
proof of the magnetic field of the sun seemed to make necessary the as-
sumption that the whole surface of the sun was negatively electrified to the
same degree as the regions about sun-spots. At least, no other possible
means of accounting for the magnetic field of the sun except by the rota-
tion of its negative charge has ever been proposed.

Upon retirement to the emeritus list of the Stanford faculty in 1919,
the author decided to attempt the identification of the hypothetical nega-
tive charges of the earth and the sun. As such an investigation was re-
garded as unpromising by my successor and former colleagues of the
Department of Physics, the university authorities declined to provide funds
or facilities for carrying on the work, and it became necessary to under-
take the investigation at my home in Palo Alto. Accordingly, a small
building was constructed and equipped for making the intended observa-
tions ; and to relieve the Department of Physics from all responsibility in
the premises it was named "The Terrestrial Electric Observatory," adopt-
ing the term suggested by Sir William Thomson for the phenomena of
earth-currents and atmospheric electricity.

fj BULLETIN Ol Mil rERRESTRIAL ELECTRIC OBSERVATORY

Knowing the superiority of the leading physical journals to any private
publication as a means of reaching the scientific public, tlu- papers giving
the earliest results of the investigation were offered to them. l>ut were
rejected, sometimes even contemptuously, by all the journals except
Science t>> which they were sent. It accordingl) became necessary to
issue them as private bulletins if they were to be published at all.

Previous to beginning the investigation it seemed probable thai if the
negative charge of the sun were great enough to account for the observed
magnetic field its inductive effect upon the earth mighl be susceptible of
observation. Uso, it did not seem impossible thai the tremendous currents
required to produce the magnetic fields of sun-spots might induce tempo
rary currents in the earth during their growth and subsidence or their
passage across the visible hemisphere of the sun. and in that manner
accounl for the terrestrial magnetic disturbances which often accompany
the appearance of sun-spots. So it was decided to test first for solar
electrostatic induction upon the earth.

One who has given careful attention to the phenomena of the atmos-
pheric potential gradient, the magnetic and the earth-current phenomena.
with their daily and seasonal variations, will see that these phenomena are
such as should appear on a conducting, electrified globe insulated in space
from other bodies and acted upon inductively by another similarly elec-
trified glohe related to it in space as the sun is related to the earth. In
fact, nearly all the phenomena which may he deduced from the known
laws of electrostatic induction under the conditions of charge which are
here assumed and the conditions in space which have been observed be-
tween the sun and the earth had been recognized before the present inves-
tigation was undertaken; but not one of them had been attributed to
electrostatic induction.

However, the most direct evidence of the sun's electrostatic induction
which could be deduced from the hypothesis of its enormous negative
charge, namely, the induction of a positive charge on the day side and a
negative charge on the night side of the earth, had not been observed.
Uninsulated conductors upon or within the earth had always been taken
as the zero from which electric charges have been measured, and the
question of the invariability of this zero had not been raised. So it has
come to be one of the tenets of physics that the earth, together with all
bodies upon its surface and including its atmosphere, contains an equal
number of elementary positive and negative electric charges of equal mag-
nitude; and while it has long been known that the earth carries a negative
e and thai bodies upon its surface are seldom, if ever, at the
true zero of electrification, it has been assumed that the electropositive
equivalent of this negative surface charge is in some manner diffused
throughout the atmosphere above the earth. The question of a possible

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE 7

difference in the distribution of these charges over the day and night
hemispheres of the earth seems not to have been sufficiently considered.

On a method of determining the distribution of the earth's surface
charge. — In theory, it seems a simple matter to insulate a body which has
been discharged to earth on the day side, to allow it to remain insulated
until it is carried around to the night side, and then to determine if it is
still in electrical equilibrium with the uninsulated body to which it was
originally discharged ; but even this simple experiment is beset with diffi-
culties. It is difficult to say when a conductor has been discharged to earth
entirely removed from the possible inductive influence of all electrified
bodies. The only known method of insuring the complete discharge of an
electrified conductor is to place it wholly inside of and in electrical contact
with a closed hollow conductor. Experiments have shown that the electri-
cal field of the earth is the same inside as outside a closed hollow conduc-
tor, regardless of any charge which may be on the outer surface of the
hollow conductor. Accordingly it has been assumed that any conductor
inside of and in contact with a closed hollow conductor is in a state of
"absolute electrical neutrality," that is, that it contains an exactly equal
number of positive and negative elementary electrical charges of equal
magnitude.

It is plain that if such enclosed conductor, after being discharged to
the outer hollow conductor, were insulated and allowed to remain inside,
it would always remain in the same electrical condition that it was in while
in contact with the outer conductor. This fact suggests a method of deter-
mining whether the distribution of the earth's surface charge is the same
everywhere around a given parallel of latitude. Thus, if two conductors
be placed inside the same hollow conductor and both be discharged to its
walls at the same time, they will be in electrical equilibrium with each other
and with uninsulated bodies upon the earth. If one of these conductors
be now insulated while the other remains in electrical contact with the
outer conductor, they will remain in electrical equilibrium with each other
unless the intensity of the earth's field undergoes a change. If this occurs,
the uninsulated conductor will gain or lose electrons until it is in equilib-
rium with the earth's field ; while the insulated conductor can neither
gain nor lose electrons, and hence cannot follow the change in intensity of
the earth's electric field. In this event a potential difference will be de-
veloped between the two enclosed conductors.

If under the proposed conditions a potential difference does appear
between the two enclosed conductors, it must be attributed to a change in
the electrical charge of the uninsulated conductor, and such a change must,
in turn, be attributed to a change in the distribution or the intensity of the
earth's surface charge.

To find whether such a change actually occurs, a quadrant electrometer

8 BULLETIN OF THE TERRESTRIAL ELECTRIC OKSI RVATORV

was set up inside an earthed wire cage; both pair- of quadrants were dis-
charged to the case of the instrument and to the wire cage; one pair was
then insulated and the other pair left connected to earth through the metal
case of the instrument and the wire cage; the needle was charged from
a battery which was enclosed in a grounded metal box, and one pole of the
battery was grounded to this box and the large wire' cage.

Sel up in this manner the instrument showed a regular daily deflection
of the needle, indicating that the insulated quadrants became electronegative
to the grounded quadrants by day and electropositive to them by night.
The deflection was greatly increased when the insulated quadrants were
connected to an insulated conductor of considerable capacity which was
suspended inside the grounded cage.

That the deflection of the needle was not caused by a change in the
electromotive force of the charging battery was shown by using an elec-
trometer with a quartz fiber suspension and charging the needle only once
in two or three days. It was found that there was a slight temperature
deflection of the needle when it was uncharged and all the quadrants were
removed ; the removal of this deflection is fully discussed in Volume III
of this Bulletin.

Many possible and impossible explanations of this daily deflection have
been proposed, and all have been found to be fallacious except the one here
given. That it was not due to a variation in the ionization of the air in the
cage was shown by placing a sheet of radioactive material on the pier
directly below the electrometer, thereby keeping the air in the vicinity of
the instrument highly ionized at all times. To show that it was not due to
a daily variation in the conductivity of the air inside the electrometer case,
the needle was grounded, the battery was placed between the two pairs of
quadrants, and one pole of the battery and its connected quadrants were
grounded. This gave a constant potential difference between the two
quadrant pairs while one pair and the needle were always grounded. The
deflections produced by this arrangement are shown in Volume V of this
Bulletin.

Finally it became clear that if all electric charges upon the earth are
due to a potential difference between the charged body and the earth, that
is, that we know nothing about electrical charges per se, then if the day
side of the earth becomes electropositive to the earth as a whole, all in-
sulated bodies near the earth on this side must become more electronega-
tive; and when the earth in their vicinity becomes more electronegative,
they must all become electropositive. Consequently, if the negative charge
of the sun induces a positive charge on the day side of the earth and a
negative charge on the night side, all insulated bodies near the earth must
become negatively charged by day and positively charged by night. That
this simple deduction was so long in dawning upon the author is humiliat-

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE V

ing; but the vain attempts to put it across to others tend to lessen the
humiliation.

It follows from the fact that similarly electrified bodies repel each
other that all insulated bodies near the earth must repel each other at all
times except twice a day.

Results of experimental tests. — To test this deduction, one diagonal
pair of quadrants was removed from the quadrant electrometer, the needle
and the other pair of quadrants were connected and insulated, and the
instrument was allowed to stand with the case grounded and the needle
and connected quadrants amber-insulated and doubly shielded from out-
side induction by the instrument case and the surrounding wire cage. No
charged body was in the vicinity of the wire cage, which was four feet
square and eight feet high.

Set up in this way, the instrument gave a very regular daily deflection,
greater than had been obtained by the previous methods of adjustment.
Thus a large Giinther & Tegetmeier electrometer with a light paper needle,
when adjusted for highest sensitivity, gave a daily deflection of the needle
through an arc of more than five degrees. To produce an equal deflection
by charging the needle and attached quadrants required a charge of more
than 200 volts. Accordingly there was a potential difference of 200 volts
between the positive and the negative charges induced in the instrument
by the changes in distribution of the earth's surface charge in twenty-
four hours, though the gain or loss of electrons by the instrument during
that time must have been insignificant and always in a direction to lessen
the effect.

Three electrometers of very different patterns and different sensitivi-
ties were set up in the same manner upon the same pier, and their deflec-
tions were recorded photographically upon the same sheet ; all gave similar
curves, but of very different range of magnitudes. Figure 1 (p. 10) shows
the mean daily variations for the same twenty days of two electrometers of
different construction, one a large Giinther & Tegetmeier instrument with
a silver suspension and a paper needle, the other a home-made electrometer
of the Dolezalek pattern with a phosphor-bronze suspension and a metal
needle. The larger instrument was more than three times as sensitive as
the smaller one. They stood side by side upon the same pier and their
deflections were recorded photographically upon the same sheet. The
average daily range of deflection was 77 millimeters in the case of the
larger instrument and 23 millimeters in the case of the smaller, the record
sheet being at a distance of approximately one meter.

The double curve in Figure 1 is interpreted as showing that the sur-
face charge of the earth at Palo Alto passed through its position of mean
intensity twice each day, once about 7 :00 a.m. and once at from 3 :00 to
4 :00 p.m., that it attained its maximum electropositive state at from 10.00

10 BULLETIN OF Nil. 1 I KR1 M RIAL llh rRIC OBSERVATORY

to 11:00 a.m. and its maximum electronegative condition at from 7:00 to
8:00 p.m. This interpretation agrees with curves previously made with
a charged electrometer needle which was deflected continuously in one
direction from 10:00 ur 11:00 a.m. to 7:00 or 8:00 p.m.. after which it
was deflected in the opposite direction throughout the remainder of the
twenty four hours. Similar curves, all of which arc very much alike, have
been obtained for seventeen months with the Dolezalek pattern elec-
trometer. A copy of the photographic record given by this instrument for
three successive days is shown on pages 16 and 17.

That the deflections are wholly due to a repulsion between the quad-
rant- and the needle when they are connected, or when both are insulated

FIGURE 1

»oon 3 lit.

Curves given by two insulated, uncharged quadrant electrometers from each of
which one diagonal pair of quadrants had been removed and the needle and remain-
ing pair put into electrical contact. The deflections are caused by the repulsion of
the needle by the attached quadrants.

separately, has been shown in several ways. When the needle is placed
midway hetween the two diagonal quadrant- its deflection is very small.
When it is rotated in such a manner as to lie partly within the quadrants
on one side it is deflected in one direction, and when it is turned so as to
lie partly within the quadrants on the opposite side the directions of its
daily deflection are reversed. That is. the needle is deflected farther and
farther away from the nearest quadrants from 7:00 A.M. to 10:00 or
11:00 a.m., after which it gradually moves back toward its nearest posi-
tion, which it reaches about 3:00 or 4:00 P.M. It then gradually moves
away from the quadrants until 7:00 or 8:00 P.M., after which it slowly
returns again to the position from which it started on the previous morning.
The periodicity of this daily variation has been very regular for the
seventeen months for which it has been recorded, but the range of deflec-
tion has differed considerahly from day to day. The daily range has been
much greater in summer than in winter, but just how much of this differ-

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE

11

ence is due to the relative positions of the earth and the sun is uncertain.
It is very noticeable that on cloudy or foggy days the deflections are of
smaller range than on clear days. Since in this climate the cloudy and
foggy days practically all occur in the winter season, it is difficult to say
how much of the decrease in the range of electrometer deflection is due
to this cause and how much to a true seasonal variation. No careful series

TABLE I

Monthly Means of Daily Deflections of Electrometer A for Seventeen

Months in 1928 and 1929

Month

A.M.
1

A.M.

2

A.M.
3

A.M.

4

A.M.
5

A.M.
6

A.M.

7

A.M.

8

A.M.

9

A.M.
10

A.M.
11

Noon

July

Aug

Sept

Nov

Dec

Feb.

Mar

Apr

May

June

July

Aug

Sept

Oct

Av. of

seventeen
months...

— 4.5

— 4.4

— 3.7

— 1.4
+ 0.2

— 2.0
+ 2.2

— 1.0

— 1.2

— 4.5
+ 1.2

— 2.2

— 3.0

— 0.2
+ 1.6
+ 7.5
+ 1.8

— 0.7

— 5.9

— 7.5
—10.9

— 5.4

— 0.7

— 3.2

— 4.0

— 4.6

— 5.9

— 2.2

— 5.3

— 6.0

— 2.0

— 0.6
+ 5.1

— 0.3

— 3.5

— 6.7
—10.0
—12.4

— 7.8

— 1.8

— 4.2

— 3.0

— 6.8

— 9.6

— 6.7

— 6.0

— 7.9

— 8.7

— 3.9

— 3.1
+ 3.5

— 1.0

— 5.8

— 7.7
—12.2
—13.3
—10.0

— 3.3

— 5.0

— 4.6

— 8.0
—12.2

— 7.7

— 8.3
—10.0
—11.5

— 6.4

— 4.6
+ 1.3

— 2.0

— 7.6

— 9.0
—13.9
—13.2
—11.5

— 5.5

— 6.0

— 5.6

— 8.7
—13.2

— 9.0
—11.6
—11.1
—12.8

— 7.2

— 5.7

— 1.1

— 2.4

— 8.9

—12.1
—16.0
—13.8
—12.6

— 6.4

— 6.0

— 6.1

— 9.5
—13.5
—12.1
—18.7
—17.0
—14.3
—10.1

— 7.0

— 3.5

— 3.4

—11.1

—15.5
—17.5
—17.6
—18.2

— 6.8

— 6.8

— 7.1
—11.1
—17.1
—17.7
—18.0
—21.2
—19.4
—17.2
—10.6

— 6.0

— 4.4

—13.9

— 3.4

— 9.7
—11.9
—12.1

— 9.7

— 7.8
—11.7
—11.6

— 7.2

— 8.8

— 7.7

— 9.8

— 9.2
—12.6
—15.3
—16.8

— 5.3

—10.2

+ 4.5

— 4.6

— 6.9

— 6.5

— 9.6

— 9.2
—11.3

— 4.9

— 0.5

— 4.4

— 4.3

— 5.9

— 4.2

— 7.2

— 8.0
—14.5

— 6.8

— 6.2

+ 7.2

— 1.6

— 0.7

— 2.0
+ 5.6
+ 5.5
+ 1.5

— 0.1

— 2.6

— 3.6

— 3.0

— 2.0

— 5.4

— 4.4

— 6.4

— 0.7

— 0.7

+ 7.3

— 0.1
+ 0.6

+ 7.5
+ 8.0
+ 4.1
+ 2.0

— 0.4

— 1.8

— 3.8

— 2.4

— 0.5

— 4.6

— 2.6

— 6.1
+ 1.2

+ 0.5

+ 5.9

— 0.1
+ 0.6

— 0.4
+ 4.6
+ 7.0
+ 3.1
+ 2.6

— 1.1

— 2.3

— 5.6

— 3.4

— 1.1

— 4.2

— 3.3

— 5.9
+ 1.2

-0.,

Month

P.M.
1

P.M.

2

P.M.
3

P.M.
4

P.M.
5

P.M.

6

P.M.

7

P.M.

8

P.M.

9

P.M.

10

P.M.
11

Mid-
night

July

Aug

Sept

Feb

July

Aug

Sept

Oct

Av. of

seventeen
months...

+ 2.3

— 3.8

— 2.1

— 3.0

— 2.9
+ 0.5

— 3.0
+ 0.2

— 4.0

— 6.2
—10.2

— 3.4

— 4.4

— 7.0

— 4.4
—10.2

— 2.6

— 3.8

— 5.5

— 8.2

— 8.9

— 8.9

— 6.4

— 3.5
—11.1

— 6.6
—10.0
—11.2
—14.7
—12.7
—11.1
—13.8
—13.3
—14.0

— 4.4

— 9.8

—11.2
—12.5
—12.8
—13.3

— 8.0

— 5.5

— 9.8

— 8.5
—15.4
—15.5
—18.7
—17.4
—16.0
—16.6
—13.9
—15.5

— 5.3

—12.9

— 9.9

— 9.2

— 9.7
—13.2

— 9.5

— 8.0
—11.2

— 9.8
—12.0
—13.0
—13.3
—14.5
—11.7
—14.0
—14.6
—16.7

— 6.5

—11.6

+ 4.9
+ 7.1
+ 2.9

— 4.0

— 5.2

— 9.0

— 3.4
+ 2.1
+ 3.9
+ 3.3
+ 3.6
+ 4.4
+ 4.7
+ 2.4
+ 0.7

— 2.4

+ 1.6

+10.5
+20.8
+23.1
+19.3
+11.5
+ 7.8
+ 9.3
+13.0
+19.2
+22.1
+18.8
+21.4
+24.7
+23.1
+22.2
+16.4
+ 7.1

+17.4

+ 15.9
+26.4
+29.6
+28.0
+13.4
+11.2
+21.4
+22.4
+29.1
+30.4
+30.4
+33.8
+31.0
+31.2
+24.8
+18.4
+10.3

+24.4

+17.1
+26.7
+28.9
+25.0
+10.6
+11.0
+18.1
+19.8
+28.4
+31.2
+32.2
+35.0
+30.6
+26.7
+21.8
+16.5
+ 9.2

+23.3

+13.1
+24.2
+23.6
+19.4
+ 8.7
+ 9.8
+13.0
+ 15.6
+21.4
+25.5
+26.2
+28.2
+20.7
+19.4
+16.7
+14.4
+ 6.0

+18.3

+ 7.4
+16.5
+17.1
+13.0
+ 7.3
+ 9.2
+ 9.4
+11.1
+14.1
+15.0
+18.0
+17.6
+ 13.5
+15.0
+12.6
+13.2
+ 5.4

+12.9

+ 2.3
+ 9.4
+ 8.2
+ 8.2
+ 4.0
+ 3.2
+ 6.6
+ 7.0
+ 7.0
+ 6.5
+11.8
+ 9.4
+ 8.3
+ 8.1
+ 7.4
+12.3
+ 4.4

+ 7.4

— 2.1

+ 2.6
+ 1.8
+ 3.4
+ 1.7

— 1.2
+ 5.1
+ 2.3
+ 2.4
+ 0.9
+ 6.2
+ 2.7
+ 2.1
+ 2.8
+ 4.1
+10.0
+ 3.3

+ 2.8

12

BULLI.IIN OF THE TERRESTRIAL 1 LECTRK OBSERVATORY

of records of cloudiness have been made, and such observations as have
been made seem to indicate that apart from disturbances caused by clouds
and fog there is considerable seasonal variation in the range of electrometer
deflection.

Table I (p. 1 1 ) gives the monthly means of daily electrometer deflections
for seventeen months, the instrument u^v<\ being the Dolezalek pattern elec-
trometer which gave the smaller range in Figure 1. During this period,
the month of December, 1 ( L H ». was unusually foggy for this climate, the
fogs frequently being dense and of long duration, ddie mean daily range
of deflection for the twenty-five days of December for which records were
measured was only seventeen millimeters, while the average daily range
for the seventeen months was thirty-eight millimeters. On one very foggy

TABLE II

Daily Variation ok Deflection of Needle of Electrometer A for Month

of May, 1929

\ M .

\ . M .

A.M. A.M.

A . M .

A.M.

A.M.

A.M.

A . M .

A.M.

A M.

Date

1

•-'

:; 4

ti

?

8

9

10

n

Noon

1

+ 2

— 2

- 5

— 9

-11

-14

—19

— 8

— 4

— 4

— 4

- 6

•)

+ 1

I

— 6

— 8

—10

—13

—19

— 8

— 5

— 8

-13

—19

- 2

-3

— 4

— 4

— 3

— 4

—20

- 6

— 3

— 2

— 3

— 4

\

- 5

s

— 9

— 9

—14

in

- 6

— 4

— 9

—10

— 7

■"»

6

— 6

9 9

- 7

—11

—14

— 4

— 2

— 1

— 1

— 4

6

+ 1

— 3

7 — 9

—12

-13

—26

-10

- 7

— 7

7

8

7

+ 5

+ 3

:; — 6

— 9

-13

—20

—10

- 7

- 6

- 6

- 7

8

-5

— 4

ti

— 8

—17

—20

— 9

— 6

— 5

— 4

- 6

9

— 1

5

12 -17

—21

-21

lii

- 5

— 2

- 2

- 1

— 3

in

+ -1

— 3—3

:.

—12

—15

— 3

+ 1

— 5

11

+ 2

• i

— 3 — 9

—11

—17

—24

11

— 9

- 8

- 8

— 8

12

+ 5

— 3

— 8 —14

—16

—29

-27

-17

-13

— 7

- 7

— 6

13

+ 3

— 1

r_' —17

—18

-18

11

—10

-3

+ 1

+ 1

+ 3

1 1

+ 2

— 5

—13 —17

—14

—17

1 1

— 9

- 3

— 4

— 3

— 2

15

—13

—14

—15 —15

—14

-15

—18

—11

— 2

+ 2

+ 2

it;

—13

—12

9

— 1

— 2

11

—19

— 8

— 1

— 1

17

-10

—15

—17

-18

—18

—18

—18

—15

—12

+ 2

+ 3

+ 1

18

— 8

— 9

— 9

—10

-10

-13

-15

-13

-5

- 5

+ 3

+ 5

19

— 7

— 7

— 8

- 8

-8

-9

—12

-13

-10

+ 3

+ 7+6

20

- 5

— 5

+ 1

— 5

-14

—15

1

- 8

- 5 -- 1

21

—11

—11

—10

—12

—14

—17

—17

—10

ii

+ 1

+ 3

+ 3

22

—11

—13

—14

—15

-15

—15

—17

—15

ti

+ 3

+ 3

+ 2

2::

-11

11

—12

—12

-13

11

1 1

— 8

+ 4

+ 1

+ 4

2 1

— 3

— 5

- 8

— 9

—12

—19

—18

- 9

— 2

- 2

+ 1

-1

— 3

- 5

- 7

—15

20

— 9

-5

::

— 2

- 3

— 1

— 3

-5

- 7

— 9

—17

—16

- 6

-4

- 3

— 2

— 3

27

— 1

— 2

— 3

— 4

— 7

—13

—15

— 4

-2

— 2

- 2

— 3

28

+ 1

— 2

- 3

— 3

— 7

17

21

—11

— 8

— 6 1 -— 6

— 7

29

— 4

11

-16

—10

- 8

1 1

—20

— 7

-5

1 — 4 - 4

30

— 4

- 8

—15

-20

22

2d

-15

-13

-2—1

— 2

■ '.1

+ 5

- 4

—11

—11

5

- 7

- 9

+ 2

+ 2

-1+2

+ 1

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE

13

TABLE II (Continued)

Date

1

2

3

4

5

6

7,

8

9

LO

11

12

13

II

15

16

17

18

19

20

21

22

23

24

25

26

'27

28

29

30

31

P.M.
1

— 9

— 8

— 9
—11

— 9
—12
—10

— 9

— 8
—16
—10

— 9

— 8

— 4

— 6

— 5

— 6
+ 5

— 1

+ 1

— 1

— 2
+ 4

— 6

— 7

— 7

— 7
—10

— 7

— 7
+ 3

P.M.
2

— 15

—13
—16
—19
—18
—19
—17
—16
—17
—13
—15
11

— 4
—13

— 9
—11
—13

+ 1

— 2
+ 3

— 6

— 8

— 2
—11
—13
—11
—16
—15
—14
—13
+ 4

P.M.
3

P.M. P.M.

4 5

P.M.

—21

—19

—18

—21

—18

—21

—23

—23

—16 ;

—15

—23

—20

— 9
—18

— 9
—15
—13

— 6

— 6

+ 1
—11
—12

— 9
—17
—20
—17
—18
—22
—19
—18

— 2

-17
-21
-12
-12
-12
-17
-22
-18
-17
-18
-23
-24
-14
-11
-14
-16

- 2
-10

- 3
-6
-12
-10

+ 4

— 6

+ 2
+ 7
+21
+ 3

— 2
+ 8
+ 4
—15

— 4

+ 1

— 3
+14
+ 1
+ 8
+14
+ 6
+ 8
—11
+12
+10
+ 10
+ 4

- 1

+ 8

+ 4

+ 6

+ 5

+ 2

+25
+11
+21
+26
+32
+25
+20
+27
+32
+ 6
+31
+31
+ 2
+34
+25
+25
+24
+19
+ 14
— 3
+26
+34
+22
+20
+14
+23
+23
+22
+27
+34
+10

P.M.

7

+35
+26
+25
+31
+35
+35
+28
+37
+37
+23
+39
+39
+20
+31
+40
+31
+34
+31
+23
+16
+30
+37
+26
+32
+27
+34
+29
+31

P.M.

+32
+37
+26
+32
+29
+32
+31
+32
+34
+29
+37
+41
+32
+25
+39
+30
+36
+33
+27
+32
+32
+33
+30
+32
+32
+31
+25
+33

P.M.

9

+34 ! +33
+37 I +33
+12 +10

+25
+35
+22
+29
+13
+30
+27
+25
+30
+26
+27
+33
+31
+21
+30
+24
+39
+19
+22
+28
+25
+26
+18
+23
+28
+21
+15
+28
+27
+23
+ 5

P.M.

10

+20
+26
+12
+22
+ 5
+22
+21
+15
+23
+19
+21
+27
+23
+ 8
+18
+15
+18
+ 1
+ 4
+ 9
+12
+16
+13
+13
+19
+11
+ 4
+18
+16
+16

P.M.
11

+10
+19
+ 2
+14

— 1
+14
+16
+ 5
+12
+ 11
+14
+19
+14
+ 1

— 2

— 4
+12

— 4

— 4

— 1
+ 5

— 1

— 2
+ 9
+10
+ 7
+ 3
+ 7
+ 8
+13

— 3

Mid-
night

+ 5

+ 9
+ 3
+ 6

— 3
+ 6
+11

— 3
+ 2
+ 8
+ 6
+13
+ 7
+ 3
—10
—11
+ 3

— 7

— 7

— 5

— 5

— 7

— 9
+ 2
+ 4
+ 2
+ 1
+ 4

+ 5

— 4

day, December 19, the total range was only five millimeters. This is the
smallest range yet observed on any day.

In order to show the uniformity of variation from day to day, Table II
gives the hourly deflection in millimeters from the mean for the day of the
electrometer needle as photographed upon a record sheet at a distance of
one meter from the instrument. This month is chosen because the daily
variations were approximately a mean for the year, and because it is the
only month of the seventeen for which a record was obtained for every day.

In Table III (p. 14) the same data are given for the month of December.
1929, which is the most anomalous of the seventeen months under considera-
tion.

It has been assumed that the mean potential of the earth at Palo Alto
occurs about 7 :00 a.m. and about 3 :00 or 4 :00 p.m., the time when the
electrometer needle is in its position of nearest approach to the quadrants,

14

BULLETIN Ol ill! rERRESTRIAL ELECTRIC OBS1 RVATORY

and that its deflection from this position is due to both il and the quadrants
becoming charged relative to the earth. It cannot be assumed that this

lition of mean potential is of the same magnitude at all seasons of the
year, since the negative charge of the earth must always be greater on the
hemisphere which is turned from the sun; but it has been observed that
its time of occurrence does nol vary greatly with the season. Accordingly,
the deflections of .the electrometer needle have been measured from the
mean of these two positions of no deflection. In Figure -'. the mean daily
deflections for the months of .May and December, 1929, are shown in this
manner. The continuous curve -hows the mean daily deflection for May
and the dashed curve that for Decemher.

The effeel of rainy, as distinct from foggy, days upon the electrometer
deflection may he seen in the June, 1929, records. In this month there were
eight days upon which there was some precipitation, though at no time was
there an important rainfall. In Figure 3 (p. 18) the mean daily electrotn-

TABLE 111
Daily Variation in Deflection of Electrometer A for December, 1929

Dat<>

7

8
9
in
11
12
13
1 I
15
L6
17
18
in
20.
'J I
22
23
21
25
26
27
28
29
30
31

A.M.
1

+

+

+
+

+14
+ 5
- 8
+11
+ 3
+12
+ 8
+ 8

A.M.

•2

+ 5 +

+

'.'
I

::
a
:>
•J
2
u
:.
2
2
2
ii
2
- 5
+ 4
—10

- 6

- 8
+ 4
+ 3
+10
+ 6
+ 6

A . M .

+

+

9

1

3

4
- 2
-3

— 8

— 1

— 2
-3

— 1

+ 1
+ 4
+ 3

— 9

— 8

— 6

— 3
+ 6
+ 2
+ 9
+ 6
+ 2

A.M.
4

A.M.
5

A.M.
6

— 9

— 9

— 9

— 4

— 4

— 3

— 1

— 1

+ 4

+ 5
— 2

— 1

— 2

— 5

— 2

- 9

+ 8
— 9

+ 7
— 8

— 5

— 7

— 7

— 2

- 3

-3

- 3

— 3

- 3

— 1

— 1

— 2

— 4

— 7

+ 2
+ 5
+ 2
— 4

+ 3

+ 2

— 9

— 7

— 7

— 5

- 8

- 8

- 8

+ 4
— 7

+ 4
—14

+ 1
—14

+ 1

- 1

- 3

+ 5

■ 6

- 2

+ 7
— 1
—12

+ 5
-13
-13

A.M.
7

+

— 8

— 3

— 2

— 5

— 2
3
7
7
2
3
2
a
4

-13

— 9

— 7

— 9

+ 1
—14

— 3
+ 2
—14
—11

I

A.M.

A.M.

A.M.

8

9

10

- 8

— 9

+ 4

- 3

-3

- 3

— 3

— 1

+ 4

+ 2

+ 1

+ 4

— 4

— 1

— 4

— 2

— 1

+ 7

+ 2

— 3

— 1

— 8

— 6

+ 5

— 7

— 7

— 1

- 2

- 3

— 3

::

— 2

— 1

- 2

+ 2

-10

— 4

—12

- 3

—13

7,

—13

—11

—10

— 7

— 7

— 7

+ 2

— 8

— 7

— 3

— 1

- 7

- 3

—14

-15

-3

- 8

—18

— 7

- 2

—14

— 3

-14

-14

- 2

11

- 9

— 2

A.M.
11

Noon

+ 4
+ 4
+ 8
+ 2

— 5
+ 7

— 1

+ 5
+ 6
+ 4

— 3

— 1

3
3
2
3
3
- 5

— 7

— 3

— 3
o

+

+

f

+

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE

15

TABLE III (Continued)

Date

7,
8
9
in
11
12
13
II
L5
16
17
18

1!)
20
21
22
23
24
25
26
27
28
29
30
31

P.M.

P.M.

P.M.

1

2

3

— 3

— 5

+ 4

+ 1

— 2

+ 5

+ 5

+ 4

— 1

— 5

— 3

— 2

+ 6

+ 4

+ 2

+ 2

+ 2

— 7

— 6

— 8

-1

-3

— 4

+ 6

+ 3

+ 1

+ 1

— 2

— 3

+ 4

+ 2

+ 1

+ 3

+ 4

+ 2

—10

—12

—10

— 9

—12

—12

— 6

— 9

—12

— 5

— 8

— 9

+ 2

— 1

— 6

— 7

— 9

— 6

+ 1

— 2

— 3

—13

—13

—13

— 9

—12

—16

— 9

-13

—11

—12

—13

—10

— 5

—10

—12

P.M.

4

- 3

1

-3

- 2
-3
-3

6

- 2

- 2

- 1

10
12
13

- ii

- 7

- 4
-5
-16

19
13

- 9

13

P.M.
5

+12
+ 2

— 3

— 4

— 1

— 3
+13

— 3

— 1

+ 1

— 1

— 2

— 1

— 8

— 1

— 7
+ 8

— 7
+ 2
—14
—15
—14
—13

P.M.

+19

+ 9

+ 1

— 2
+ 2

— 1
+17
+ 8
+ 7
+10
+ 1
+ 2
+ 8
+17
+12
+ 9

— 5
+ 9
+10
+21
+19
+13

— 9

P.M.

7

+21
+17
+ 1

+ 3
+ 1
+ 1
+15
+10
+11
+10
+ 1
+ 4
+10
+16
+13
+13
+ 4
+ 8
+ 5
+21
+24
+16
+17
+16

P.M.

+14
+ 9

— 2

— 1
+ 4

— 2

+13
+16
+ 9
+ 6

+ 2
+10
+ 9
+14
+14
+11
+13
+ 5
+16
+18
+13
+18
+21

P.M.

9

+

+

+15
+ 3

— 3

— 1
+10

— 2

- 3
5
3
2
1

- 1

+11
+ 8
+12
+12
+13
+12
+ 7
+15
+15
— 8
+17
+18

P.M.
10

+

9

1
3
1
6

— 2

— 2

— 5

— 3

— 1

— 1

+10
+ 9
+10
+13
+13
+10
+ 5
+15
+13

+15
+18

P.M.
11

+

9

1

3
1
6

- 2

- 3

+10

— 3

— 1

+ 5

+11
+12
+14
+ 6
+ 6
+17
+10
+12
+13
+13

Mid-
night

9
3
3

t;

2
3
2
9
2
2
1

5
- 6
+ 9
+13
+13
+ 7
— 6
+14
+ 7
+14
+12
+ 9

+

+

eter variation for the eight days upon which there was observable precipi-
tation is shown by the dashed curve, and the mean variation for eighteen
days without precipitation is shown by the continuous curve.

Relative magnitudes of the deflections of the needle of Electrometer A for the
months of May and December, 1929. The continuous curve shows the mean daily
range of deflection for May, and the dashed curve the range for December.

Lunar electrostatic induction. — In previous volumes of this Bulletin
attention has been called to the effect of the moon's electrostatic induction

16

ETIN Ol iiii rERRESl RIAL ELECT RIC OBS1 R\ A fORV

10 11 KCOK 1

Copy of photographic record for three successive day- ol
and the remaining pair and needle connected, uncharged and
the copy three-fifths of the original.

THE EARTH'S ELECTRIC CHARGE AND THE BAROMETRIC WAVE 17

***

"V/v*

11

MT.

actions of electrometer A with one pair of quadrants removed
ated inside a metal case in a grounded wire cage. Scale of

18 BULLETIN <>!•" THE TERRESTRIAL ELECTRIC OBSERVATORY

FIGURE 3

The continuous curve shows the mean daily electrometer deflection for eighteen

days without precipitation, and the dashed curve shows the mean deflection of the
same instrument for eight days upon which there was noticeable precipitation, both
in June, 1929.

upon the earth. This effect is also shown by the present arrangement of

apparatus: hut less plainly than by the earlier methods, ami the daily range
of deflection due to the moon's induction seems to bear a smaller ratio to
that of the sun than it has in past experiments when a charged electrometer
needle was used.

In order to test for the lunar effect, the deflections which have been
measured at hourly periods throughout the solar day must he distributed
according to lunar hours. For this purpose the measurement which was
made nearest the time of the moon's upper culmination is taken as the mid-
hour of the lunar day. The time of this measurement may vary as much
as thirty minutes on either side of the time of upper culmination. Then,
unless every hour of a synodical period is used in the determination of the
mean daily deflection, a considerable error is introduced, due to the much
greater magnitude of the solar than of the lunar deflection. Unfortunately,
it was difficult to find complete synodic periods without any missing days.

In Figure 4 the continuous line gives the mean lunar diurnal variation

FIGURE 4

Lunar diurnal variation of Electrometer A. The continuous curve shows the
mean lunar daily electrometer deflection for three synodic periods, and the dashed
curve shows the mean deflection tor one hundred and ninety-six lunar days taken
without regard to synodic periods.

for three synodic periods, and the dashed curve gives the mean lunar diur-
nal variation for one hundred and ninety-six days taken without reference

THE EARTH'S ELECTRIC CHARGE AND THE BAROMETRIC WAVE 19

to synodic periods. The mean range of daily variation for the continuous
curve is four and three-tenths millimeters.

Some bearings of the observations upon electrical theory. — It seems
that the observations described above are conclusive as to the electrostatic
induction of the sun upon the earth, and that no more direct proof of the
electrical charge of the sun can be hoped for. They likewise show con-
clusively that zvhat we call electrification is a condition depending upon the
electrical state of the body relative to the earth, and that we know nothing
whatever regarding a condition known as "absolute electrical neutrality."
Our modern atomic theories which are based upon the assumption of per-
fect equality in the number and magnitude of two kinds of elementary
electric particles in every complete atom has no basis in experimental fact.
An atom which is "neutral" upon the earth might be a highly charged atom
upon some other planet or sun, and vice versa. Attention has been called
in earlier volumes of this Bulletin* to evidence that atoms which are highly
charged upon the earth may be uncharged in some of the stars and nebulae,
and even upon our own sun. It is the hope of the author to discuss these
matters further in a monograph on the subject of "Terrestrial Electricity,"
upon which considerable work has already been done.

Daily variation in atmospheric pressure and the distribution of the
earth's surface charge. — In a paper published in Science for April 19, 1929,
the question is raised as to a possible influence of the variation of the
earth's surface charge upon atmospheric pressure. We have seen that one
of the necessary conclusions resulting from the observations described in
the foregoing pages is that every body insulated from the earth in lower
or middle latitudes necessarily becomes electrically charged twice each day,
once positively and once negatively. This statement must apply to the gases
of the air as well as to small particles floating in the atmosphere.

The charges acquired by the molecules of the air do not involve any
physical change in them, but result wholly from the changes which take
place in the earth's electrostatic field. Hence, without any change in their
electron or proton content, every so-called neutral molecule in the earth's
atmosphere must be attracted toward the earth twice daily. Also, every
such molecule in the earth's atmosphere must repel every other similar
molecule in its vicinity twice each day. Can these conclusions be further
verified by observing any results of the molecular charges?

There are two phenomena in connection with the daily variation of
atmospheric pressure which have never been satisfactorily explained ;
namely, the times of occurrence of the morning and evening maxima and
the cause of the twelve-hour pressure wave. Both phenomena are very
uniform over the earth. Regarding this uniformity, Hann tells us :

* Volume IV, page 22.

20 BULLETIN OF THE TERRESTRIAL ELECTRIC OBSERVATORY

No other meteorological element has so regular a daily period as the atmospheric
pressure; and this in spite of the fact that the amplitude of this daily variation ia
relatively small, ranging from two or three millimeters in the tropics to a few tenths

of a millimeter at 60 latitude. The daily period is double; the atmospheric pressure
reaches twice daily a maximum ami twice a minimum, and, where the daily atmos-
pheric pressure is least disturbed, both maxima and minima are very much alike. This
IS very different from the daily range of other meteorological elements, and suggests
tlv ebb and flow of the sea, for which reason these waves have been called atmos-
pheric tides. In spite of their resemblance in form, an important difference in the
two phenomena appears in that the atmospheric ebb and flow follows the sun and
occurs according to true local time, and that no lunar influence is perceptible in it.
Accordingly it can not be a gravitation phenomenon, since in that case the lunar
period would be much more strongly marked than that of the sun.

The phenomenon has, accordingly, a much greater theoretical interest than the
daily periods of the other meteorological elements, which, although much less simple
and locally much more variable, yet can be definitely shown to depend upon the con-
ditions of insolation. Practically, on the contrary, the daily barometric variation, on
account of its minuteness, is of little significance and can scarcely be related to any
consequences, while the daily period of temperature, for example, is regarded as of
great importance and occupies a very conspicuous place in the domain of meteorology.*

A remarkable characteristic of the semi-diurnal barometric variation
is the regularity of the occurrence of the maxima and minima and their
uniformity in time of day in all latitudes. While the amplitude of these
waves may vary greatly with latitude, with elevation, and with location,
whether over the sea or over the land, the local times of maxima and
minima are very constant. This is true also for the different periods of
the year, though the amplitude of variation is everywhere greatest at the
equinoxes and least at the solstices.

In this regular daily variation of atmospheric pressure there is a fore-
noon maximum which occurs quite uniformly from nine to ten o'clock
wherever it has been recorded. This is a time of day when we should
expect a low, instead of a high, barometer. It is customary to attribute a
falling barometer over a given region to vertical atmospheric convection
caused by the heating of the air over that region. Since the temperature of
the lower air depends principally upon that of the surface of the earth
beneath it, we look for a low barometer over the warmer regions of the
earth.

The lowest temperature of the air over the land occurs about four
o'clock in the morning and over the sea it occurs about midnight, or a
little later. From this time forward the temperature of the air rises until
nine or ten o'clock in the forenoon. Both the temperature and the baro-
metric pressure rise most rapidly about seven or eight o'clock in the morn-
ing. During this time convection becomes well established. So it follows
that over regions where the atmosphere near the surface is lighter the baro-
metric pressure is greater than over surrounding regions where the surface

Hann, Lehrbuch drr Meteorol 177. Translation by the present writer.

THE EARTH'S ELECTRIC CHARGE AND THE BAROMETRIC WAVE 21

atmosphere is heavier. We know it is heavier, because it is displacing the
air over the regions where the barometer stands higher, otherwise there
could be no convection over these regions.

One attempt to explain this paradox which had the approval of a num-
ber of well-known meteorologists seems to have been first proposed by
Espy, in 1840. It is based upon the assumption of a manometric effect of
the heated air near the ground, which was supposed to be hindered in its
expansion by the heavier air above it and could not, for that reason, dis-
tribute its pressure to the higher air, and hence caused the barometer to
register a higher pressure than it would if the compression could be dis-
tributed to the whole vertical column of air. This seems to assume that
the air resting upon the heated volume below must all be forced upward
as a rigid body, and that it does not diffuse into the surrounding air as a
result of its compression but can only spread out at the top of the atmos-
phere. Then, while the expanding volume below is giving an acceleration
against gravity to the cooler mass above it, it is supposed to react upon the
earth and thus increase the barometric pressure.

Now it is well known that the distinguishing characteristic of an elastic
body is that it reacts to a stress in such a manner as to distribute the stress
uniformly throughout the whole body. The rate at which a stress is dis-
tributed is the rate at which an elastic wave will travel through the medium.
In the case of a gaseous body, the rate at which a compression will be
distributed throughout the whole volume is the rate at which a compres-
sional wave will travel through the gas. This rate for an isothermal com-

E
pression is given by the equation V 2 = — . Accordingly, a compression

set up in our atmosphere, even at a temperature of — 20° C, will be dis-
tributed throughout the whole atmosphere at a speed of a little more than
eight hundred feet in a second, more than nine miles in a minute.

At a height of twenty-four miles above the earth the atmospheric pres-
sure is only one six-hundredth as great as it is at the ground. A compres-
sion at the ground would be distributed through a vertical column twenty-
four miles high in two and seven-tenths minutes, even at the low tempera-
ture specified.

The rate of temperature change in the lower atmosphere is, in equa-
torial regions, generally less than two degrees an hour. If a complete
vertical column of air were confined and heated at this rate, its pressure
would increase by about five millimeters an hour, or by about one-fourth
of a millimeter in the time which would be required for the pressure to be
uniformly distributed from the lower end of the column.

Untenable though the proposed explanation is seen to be, it crops out
in a slightly modified form in our most important American textbook on
Meteorology. In this case, it is assumed that the surface winds on the

22

BULLETIN OF THE TERRESTRIAL l LLC1R1C OHSKRVAToRY

earth arc dammed up whenever they flow into a region of vertical convec-
tion, and thus ^et up an excessive barometric pressure due to their decrease
speed. Thus the author >a .

It is obvious that the more- active vertical convection becomes, the greater will
be its interference with the flow of tin- atmosphere, the more winds will be dammed
up and the higher the barometric pressure. As convection increases, reaches a maxi-
mum, and then decreases, so, too, will the resulting interference go through the same

changes.*

As a proof of this "obvious" statement, the author shows that if air
which is moving slowly over the earth should rise and be replaced by
higher air which is moving more rapidly in the same direction, and if the
colder air after settling down to earth should have its velocity decreased
by ground friction, the resultant velocity of the whole volume of air under
consideration would be less than it was before. This is supposed to slow
down the surface air which was following, and to cause it to be compressed,
although from the premises the exchange of positions of the slower volume
below with the faster volume above should increase the velocity of the
surface winds.

It is not profitable to spend time on the discussion of these explanations
of the ten o'clock barometric maximum. Both <>i them are based upon the
assumption that this maximum is, in some manner, brought about by the

FIGURE 5

Noon 6

\

/
/

/
i

i

i \

/
/
/

/

i

\

\

\
\

\

*~ \

A

\ _

The upper pair of curves show the mean daily range of temperature and baro-
metric pressure at Toronto for the month of January, Y)ll , and the lower pair show
the same data for the month of July, 1927. The continuous curves represent the
barometric variation and the dashed curves the temperature.

* Humphreys, Physics of the Air, page 236.

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE

23

rapid convection of the air at this time of day, while it is possible to show
that this particular barometric maximum is wholly independent of con-
vection and is, in fact, most plainly marked in regions where convection
is least.

For example, the Meteorological Service of Canada has published
monthly statements of the hourly variations of barometric pressure and of
temperature for every day in the year and for a number of stations. The
curves in Figure 5 give these data for the months of January and July,
1927, at Toronto. The upper pair of curves show the mean daily variation
of barometric pressure and temperature for thirty-one days in January,
and the lower pair of curves show the same data for thirty-one days in
July. Both sets of curves are drawn to the same scale. The continuous
curves show the barometric pressure and the dashed curves show the tem-
perature. The mean daily temperature for January was twenty-two degrees
Fahrenheit, and for July it was sixty-eight degrees.

It will be seen from these data that while the mean daily range of
temperature was nearly three times as great in July as in January, the ten
o'clock barometric maximum was higher in January than in July. Half
the forenoon barometric rise in both months occurred while the tempera-
ture was at the lowest point for the day, and, consequently, before convec-
tion could have begun. The same phenomenon is shown in the data for
the coldest day in January, when the highest barometric pressure for the
day occurred at ten o'clock in the morning while the temperature was six
degrees below zero.

On page 61 of Hann's Lehrbuch der Meteorologie are given tables
showing the mean daily variations of summer temperature of the water
and the air at 0°-10° N. and at 30° N. over the Atlantic, as taken from the
records of the Challenger Expedition. In Figure 6, the curves for daily
variation of water and air temperatures at 4?5N. are shown. The vertical
line drawn at two o'clock p.m. represents one degree Centigrade, showing
that the maximum difference in temperature of the air and water was
approximately one degree, the water being always warmer than the air.
Since this maximum temperature difference occurred at four o'clock in

FIGURE 6

Noon

U.

Mean daily relation of air temperature to water temperature over the ocean at
4?5 N. latitude. The vertical line at 2:00 p.m. represents one degree Centigrade.

24

BULLETIN "1 THI rERRESTRIAL ELECTRIC OBSERVATORY

the morning, whatever convection there was .sin mid have been greatest
at this time.

The same temperature difference between the air and the water is
shown in a different manner in Figure 7, the air temperature being taken

FIGURE 7

Kxcess of water temperature over air temperature compared with the daily range
of barometric pressure over the ocean at 4°5 N. The continuous curve indicates the
temperature and the dashed curve, the barometric variation.

as the base line from which the water temperature is measured. The same
scale is used as in Figure 6. The dashed curve in Figure 7 shows the
daily variation in barometric pressure in summer over the same region,
taken from page 178 of Hann's Lehrbuch. It will be seen that the baro-
metric pressure has no appreciable relation to whatever convection may
be caused by the temperature difference between the air and the water.

A comparison of the barometric data over this region and over inland
regions farther north shows that the morning barometric maximum is as
high over the ocean in equatorial regions as it is over the land where the
daily temperature variation is ten times as great and where the maximum
temperature for the day occurs more than twelve hours later. These facts
seem to suggest some other agency than convection as the cause of the
morning barometric maximum and of the twelve-hour variation in baro-
metric pressure as observed over the ocean in equatorial regions.

1 1 aim says :

The daily range of temperature of the air over the ocean is practically independent
of temperature changes in the surface of the water. The air can give off no heat to
the water at night which it has absorbed from the water during the day. The daily
range of air temperature over the ocean must depend, in the main, upon the absorp-
tion of the sun's rays and the radiation to the sky. It is easily to be seen that under
these conditions the temperature range must be very small.*

Convection depends upon the heating of the air by the earth below it.
If there is any convection over the ocean, it must be principally in the
early morning hours, and the low barometer which occurs at this time
suggests such a possibility. There seems no reason to doubt that the
vertical convection caused by heating the air near the ground does cause
a decrease in barometric pressure ; hence there must be a daily variation

* Loc. cit., page 61.

THE EARTH'S ELECTRIC CHARGE AND THE BAROMETRIC WAVE 25

in barometric pressure due to the daily change in temperature of the air
near the ground. Such a variation should produce a twenty-four-hour
wave, and not a twelve-hour wave such as is observed over the ocean in
low latitudes. This fact has long been recognized, and there have been
various attempts to separate the twenty-four-hour convectional wave from
the total barometric wave for the day. These attempts have hitherto been
unsuccessful. Hann says, in discussing the two systems of waves :

The whole day wave is subject to very great local disturbances, so that it is very
difficult to separate the universal terrestrial remainder of the same from these dis-
turbances

The half-daily wave is the principal phenomenon, and has a very regular course,
such as is not to be found in any other meteorological phenomenon. For this reason,
an unknown cosmical cause has sometimes been assumed as its explanation.*

The twelve-hour barometric wave is, then, a hitherto unexplained phe-
nomenon. The one attempted explanation which has acquired some stand-
ing among meteorologists is based upon a surmise by Lord Kelvin that the
atmosphere as a whole may have a natural twelve-hour period of oscilla-
tion, and that this might cause the twelve-hour wave to be set up by the
twenty-four-hour wave. A number of learned mathematical papers have
been written on this subject, and some meteorologists seem to find a satis-
factory explanation in it ; but it would seem not to require any excursion
into higher mathematics to decide that the twelve-hour barometric fluctu-
ation is not due to a natural oscillation of the atmosphere. In the first
place, such an oscillation would have fixed nodes ninety degrees apart upon
the earth. The atmosphere cannot oscillate as an elastic sphere, since it
forms a very thin compressible skin over the surface of an incompressible
globe. Any compression set up in a part of this elastic layer must travel
around the earth as a compressional wave at a speed of less than ten
miles a minute, and would require more than forty hours to travel around
the earth at the Equator.

Again, a compressional wave would not follow the parallels of latitude,
but would spread out in all directions with equal speed so long as it was
in a region of uniform temperature ; hence places on the same meridian
would not be in the region of maximum pressure at the same time. As it
is, all places on a given meridian from one polar circle to the other have
maximum barometric pressure at the same actual time, and a similar pres-
sure belt extends in the same way along the meridian 180° distant. The
barometric pressure belt, whatever its cause, extends entirely around the
earth in a meridional direction. But it does not rotate with the earth, but
remains fixed relative to the sun, while the earth rotates under it. As the
earth rotates from west to east the barometric pressure wave moves around
it from east to west.

* Loc. cit., page 192.

26

BULLETIN OF THE fERRESTRIAL ELECTRIC OBSERVATORY

Although the twelve-hour wave and the twenty four-hour wave have not
hitherto ! irated, it' it be true that we have a region over the ocean

where the barometric variations are independent of convection, it would
seem that the inland convectional wave in the same latitude might he de-
termined by subtracting the oceanic barometric wave from the total baro-
metric wave inland. Unfortunately, land and sea curves for barometric
variation in lower latitudes are not at hand at the time of this writing.
ll:mn gives data on the barometric variation over the ocean at 33 X. and
at Zurich. Switzerland, which is more than ten degrees farther north. If
ceanic curve he subtracted from the Zurich curve we should have
[( ft a curve approximating that due to convection at Zurich. The resulting
curve is a twenty-four-hour curve, showing a maximum barometric pres
>ure at 4:00 A.M., the coldest part of the day, and a minimum pressure at
6:00 p.m. The maximum rate of decrease of pressure occurs between
6:00 A.M. and 8:00 A.M., the period of maximum rate of increase of
temperature.

Hann also gives data on two other pairs of curves which may be used
to throw light on our problem. In order to compare inland and oceanic
stations in nearly the same latitude he selects the Island of Jersey and the
station at Kalocsa on the Danube Plain in Hungary. When the curve
showing the barometric variation at Jersey is subtracted from the Kalocsa
curve, a twenty-four-hour curve is obtained similar to the one obtained
by subtracting the ocean curve from the Zurich curve, the maximum and
minimum pressures occurring at the same time in both cases.

Another curve showing the difference between inland and coast stations
may be had by subtracting the curve showing the daily variation of
barometric pressure at Yalentia Island, on the coast of Ireland, from the
corresponding curve at Greenwich.

The three twenty-four-hour curves formed in this manner are shown
in Figure 8. The continuous curve represents the Zurich-ocean curve, the

Three twenty-four-hour barometric curves derived by subtracting oceanic
ctcr curve- from inland curves.

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE

27

long-dashed curve represents the Kalocsa-Jersey curve, and the short-
dashed curve the one for Greenwich-Valentia. Each probably represents
with some degree of approximation the twenty-four-hour convection curve
at the respective inland station, but none of them is probably a very accu-
rate representation of this curve. However, it seems probable that if true
oceanic curves could be compared with inland curves for the same latitude
a close approximation to the true convection curves could be obtained.

Attempted explanation of the twelve-hour barometric wave. — It seems
evident that the inland barometric wave consists of a twenty-four-hour
wave due to convection and a twelve-hour wave which is independent of
convection. This twelve-hour wave is hitherto unexplained, and the main
purpose of the present discussion is to inquire if it may be due to the
semi-diurnal attraction by the earth's charge of the air molecules and the
floating particles of water and dust in the air, combined, perhaps, with the
repulsion of these particles for each other.

The daily electrometer and barometer curves do not closely resemble
each other, as may be seen from Figure 9, in which the continuous curve

FIGURE 9

Comparison of electrometer and barometer curves at Palo Alto for March, 1929.
The continuous curve shows the electrometer deflection, and the dashed curve shows
the mean daily barometer range for the same days.

represents the diurnal electrometer deflections and the dashed curve shows
the daily barometer variations at Palo Alto for the same period, March,
1929. However, when it is recalled that the electrometer deflections indi-
cate a decrease of the mean surface charge of the earth between 8 : 00 a.m.
and 4 : 00 p.m., it will be seen that during this time the barometer is falling,
while it is rising during the whole time of increase of the earth's negative
charge. This seems to indicate that the atmosphere near the earth is, on
the whole, electropositive to the earth, a fact which has long been known.
When the earth becomes less electronegative and the atmosphere becomes
less electropositive, the barometric pressure is correspondingly decreased.
In Figure 10 (p. 28) the continuous curve gives the mean of the three
twenty-four-hour curves shown in Figure 8 and the dashed curve shows a
twenty-four-hour curve given by subtracting the mean daily electrometer
curve for five months from the mean daily barometer curve at Palo Alto
for the same period.

28

BULLETIN OF THE TERRESTRIAL ELECTRIC OBSERVATORY

FIGURE 10

:.'oon

Comparison of the mean of the three twentj four-hour barometer curves of Fig-
ure 8 with a curve derived by subtracting the mean daily electrometer curve for live
months from the mean daily barometer curve for the same period. The continuous
curve is the mean of the three curves of Figure 8, and the dashed curve is the Palo
Alto curve.

There are, no doubt, many other resemblances between the daily elec-
trometer curve and the twelve-hour barometer curve which will be ob-
served when the two curves are better known. On page 234 of Physics
of the Air, Humphreys mentions a number of characteristics of the semi-
diurnal barometric variation. Among them is the fact that the barometric
variations are less on cloudy days than on clear days, which we have also
seen to be true of the electrometer variations. Another statement is that
the amplitude of the barometric variations is everywhere greatest at the
equinoxes and least at the solstices. That the same is true of the amplitude
of the earth-potential variations was shown in Volume V of this Bulletin.
In Figure 11 the curve there given for the monthly mean daily ranges of

FIGURE 11

J 1 M L U J J k 3 N 9

/

v

;

i
i

^

\

\\

i

\

//

7

^-

\v

\

\
\

\

'/

/

V

\
\

A

1

The monthly mean daily range of electrometer deflections for 1927 compared
with two turves showing the monthly ranges of daily barometer variation for sta-
tions in Europe. The continuous curve shows the electrometer deflections.

electrometer deviations for the year 1 ( >27 is compared with two curves for
monthly range of daily barometric variations. The continuous curve is

THE EARTH S ELECTRIC CHARGE AND THE BAROMETRIC WAVE

29

for the monthly mean daily range of electrometer deflection for 1927, the
long-dashed curve is the monthly range of the twelve-hour barometric
wave for the four stations, Milan, Turin, Modena, and Rome, as given by
Hann (page 190), and the short-dashed curve is the mean daily range of
barometric pressure for the different months for eight stations in Europe,
as given by Arrhenius, Kosmische Physik, page 603.

Relation of barometric wave to earth-currents. — In the Bidlingmaier
diagram of Figure 12 the continuous curve represents the mean continuous

FIGURE 12

The continuous curve represents the mean continuous distribution of atmospheric
pressure around the earth in equatorial latitudes, the long-dashed curve represents the
mean intensity of the W-E earth-current around the parallel of Berlin, and the short-
dashed curve represents the mean intensity of the total resultant earth-current around
the parallel of Tortosa, Spain.

distribution of barometric pressure around the earth in equatorial regions,
the long-dashed curve represents the intensity of the W-E earth-current
around the parallel of Berlin for four years, as given by Weinstein,* and
the short-dashed curve shows the mean distribution of intensity of the
total resultant earth-current at Tortosa, Spain, for the years 1910-26, as
given by Puig.f The resemblance of the three curves seems close enough
to indicate a physical relation between the phenomena which they repre-
sent.

Atmospheric potential gradient and barometric pressure. — Resem-
blances between the daily variations in barometric pressure and the atmos-
pheric potential gradient have frequently been observed, and attempts have

* Die Erdstrome , Tafel 1.

t Puig, Las corrientes teluricas en Tortosa, page 72.

30

BULLETIN OF Mil. fERRESl KI AI. l-.I.H'TKIC UBSI-RVATiUn

been made to explain the potential gradient variations l>y the movements
of atmospheric ions caused by vertical convection. The comparisons which
have hitherto been made between these two phenomena have always taken
into consideration the total barometric variation, which was assumed as
due to convection.

It has been shown in the preceding pages that the convectional baro-
metric variation gives a twenty-four-hour wave, while it is well known
that at low altitudes the potential gradient variation gives a double wave in
twenty- four hours. This would seem to indicate that if the potential gradi-
ent variation is related to either of the barometric waves it must he to the
twelve-hour wave, which is not due to convection.

This is shown very clearly when the two barometer waves are compared
with the potential gradient wave. In Figure 13 the continuous curve shows

FIGURE 13

Comparison of the daily variation of air-potential gradient at Kew with the
twelve-hour and twenty-four-hour barometric curves. The continuous curve repre-
sents the mean daily variation of air-potential gradient at Kew. The short-dashed
curve gives the mean daily variation of barometric pressure at Jersey, and the long-
dashed curve shows the twenty-four-hour curve derived by subtracting the Jersey
barometric curve from the barometric curve at Kew.

the mean daily variation in atmospheric potential gradient at Kew* for the
year. The nearest ocean station to Kew for which data are at hand is the
Uland of Jersey. The short-dashed line shows the mean daily variation of
the barometer at Jersey, and the Ion- dashed line shows the convectional
barometer wave at Kew as nearly as it can be determined by subtracting
the Jersey curve from the total barometer curve at Kew. Attention h
called to the agreement of this curve with the three convectional barometer
curves shown in Figure 8. It is readily seen that the potential gradient
curve hears no appreciable relation to the convectional barometric curve.

* Mache und von Schweidler, Die Atmospharische Elektrisitat, pag<

THE EARTH'S ELECTRIC CHARGE AND THE BAROMETRIC WAVE 31

On the other hand, it bears much the same relation to the twelve-hour
Jersey curve that the latter does to the earth-potential curve. The relation
of the earth-potential and air-potential gradient curves has been considered
on pages 16-17 of Volume II and on page 17 of Volume III of this
Bulletin.

Funds for the publication of this volume have been provided by the
Committee on University Publications, Stanford University.